Retrograde Michael Reaction in Additions of Active Methylene

Retrograde Michael Reaction in Additions of Active Methylene Compounds to Tetracyanoethylene1. J. W. Van Dyke Jr., H. R. Snyder. J. Org. Chem. , 1962,...
0 downloads 0 Views 333KB Size
VAS DYKE,JR.,AND SKYDER

3588

Retrograde Michael Reaction in Additions of Active Methylene Compounds to Tetracyanoethylene’ J. W. VANDYKE,J R . ,ASD ~ H. R. SNYDER iVoyes Chemical Laboratory, University of Illinois, Urbana, Illinois Received April 5, 1968 Tetracyanoethylene reacts with 2H,3H-thieno [3,2-b]pyrrol-3-one and 2H,3H-benzo [b]thiophene-3-one to give 2-dicyanomethylene-2H,3H-thieno[3,2-b]pyrrol-3-one(11) and 2-dicyanomethylene-2H,3H-benzo[b]thiophene-3-one (VI), respectively. Compounds I1 and VI undergo Diels-Alder reactions giving adducts containing spiro systems.

Tetracyanoethylene (TCNE) has been reported to react with aromatic or heterocyclic nuclei to introduce a tricyanovinyl group int,o the ring s y ~ t e m . ~ The - ~ mechanism proposed for this reaction involves initial addition across the double bond in tetracyanoethylene, followed by the base-induced elimination of hydrogen cyanide (equation 1). The tricyanovinyl group can be converted to a carbethoxyl group by a series of solvolytic steps.3

Ketones possessing an cY-hydrogen react with tetracyanoethylene to yield substituted tetracyanoethanes (equation 2).6 The reaction is very slow when no catalyst is used but can be accelerated by 0

I’ RCCHR’R’

+ C(CN)%-+ &(CN), 0

ylene would follow one of the above patterns; reaction according to scheme (1) would afford a means of introduction of the carbethoxyl group into the pyrrole ring of the heterocyclic ketone. When I was refluxed with tetracyanoethylene in dry benzene, a deep maroon solution resulted. After all solid had dissolved, a small quantity of pyridine was added and a maroon solid precipitated. The n.m.1. spectrum possessed doublets at 7 = 2.35 p.p.m. ( J = 2.4 c.P.s.) and 7 = 3.56 p.p.m. ( J = 2.4 c.P.s.), consistent with the presence of hydrogen in the a- and &position of the pyrrole ring, respectively. The product was, therefore, not a tricyanovinyl derivative; the n.m.r. spectrum, the composition, and the reactions mentioned below indicated it t o be 2dicyanomethylene-2H,3H-thieno [3,2-b]pyrrol-3-one (11).

QTN)2 H

I1

The course of the reaction can be explained by a mechanism similar to the formation of the tricyanovinyl derivatives. After the addition across the double bond of tetracyanoethylene, instead of abstraction of the terminal hydrogen by the base,

C H ( C N ) ~ - C ( C N ) ~ - R ’ R ~ ~ R(2)

addition of the “molecular” silver catalyst of Gomberg and Cone7 or by volatile acid catalysts, such as hydrogen chloride or boron trifluoride. The adducts with ketones, however, were found to be unstable to basic catalysts and elevated temperR

Results and Discussion It was to be expected that the reaction of 2H,3Hthieno [3,2-b]-pyrrol-3-one (I) with tetracyanoeth(1) Supported in part by 8 research grant [C3969-B1o] from t h e National Cancer Institute. Public Health Service. (2) National Science Foundation Cooperative Fellow, 1960-1961. (3) B. C. McKusick, R. E . Heckert, T. L. Cairns, D. D. Coffman, and H. F. Mower, J. A m . Chem. Soc., 80, 2806 (1958). (4) G. N. Sausen, V. A. Engelhardt, and W. J. Middleton, %bid., 80, 2815 (1958). (5) W. E . Noland, W. C. Kuryla. and R. F. Lange, abid., 81, 6010 (1959). ( 6 ) W.J. Middleton, R. E . Heckert, E . L. Little, and C. G Kreapan, i b z d . . 80, 2783 (1958). (7) M. Gomberg and M. Cone, Ber,, 89, 3286 (1906).

the ring hydrogen is abstracted giving rise to the anion 111. The loss of the malononitrile anion gives the product I1 (equation 3 ) . The presence of malononitrile was detected b y the use of thin layer chromatography on silica gel

NOVEMBER, 1962

RETROGRADE J I I C H A E L

with dioxane as solvent. The solvent was allowed ta travel 10 cm. After being dried, the plate was sprayed with basic potassium permanganate. A brown spot in the reaction mixture corresponded to a control of malononitrile ( R f = 0.67). The reaction between I and tetracyanoethylene was repeated, but this time benzaldehyde was introduced immediately after the addition of pyridine. After I1 was removed, the reaction mixture was concentrated to dryness and benaylidenemalononitrile (IV) was isolated.

c( c N ) ~

&CH= IV

When 2Hj3H-benzo [b]thiophene-3-one (V) was treated with tetracyanoethylene, 2-dicyanomethylene-2H,3H-beneo [b]thiophene-3-one (VI) was abtained.

v

VI

Since I1 and VI contain reactive carbon-carbon double bonds, it was of interest to examine the Diels-Alder reaction with 2,3-dimethylbutadiene. When I1 and VI were refluxed with the diene for forty-eight hours in ethanol the Diels-Alder adducts VI1 and VIII, respectively, were obtained. CHB

CH,

" VI1

CHj CHj VI11

The n.m.r. spectrum of VI1 showed doublets a t = 2.35 p.p.m. (J = 2.42 c.P.s.) and r = 3.68 p.p.m. ( J = 2.42 c.P.s.) in good agreement for aand @-pyrrole hydrogen, respectively. One of the methylene groups showed a singlet a t r = 6.90 p.p.m. The other methylene group, having nonequivalent hydrogens, was split into two doublets; this probably is the methylene group adjacent to the dicyanomethylene group. The two doublets were found a t r = 6.89 and 7.57 p.p.m. ( J = 19 c.P.s.). The relative area of the outer peak to that of the inner peak was found to be 0.44 (calcd. 0.41) .8 The two methyl groups showed a singlet a t r = 8.24p.p.m. The n.m.r. spectrum of VI11 also showed similar splitting. The aromatic hydrogen absorption was centered a t r = 2.39 p.p.m. One methylene group had a singlet a t r = 6.98 p.p.m. and the two methyl groups had a singlet at r = 8.23 p.p.m. The other T

(8) L. M. Jackman, "Applications of Nuclear Magnetic Resonance Spectroscopy in Organic Chemistry." Pereamon Press, Inc., New York, N . Y . . 1959, pp. 89-90.

REACTIOS

3889

methylene group was again split into two doublets at r = 6.86 and 7.66 p.p.m. ( J = 17 c.P.s.). The relative area of the outer peak to that of the inner peak was equal t o 0.50 (calcd. 0.50).8

Experimentalg 2-Dicyanomethylene-2H,3H-thieno[3,2-b]pyrrol-3-one (II).-To a boiling solution of tetracyanoethylenelo (0.46 g.) in 25 ml. of dry benzene was added 0.5 g. of 2H23H-thieno[3,2-b]pyrrol-3-one. The mixture was heated a t reflux for 10 min. (or until all solid dissolved) forming a dark maroon solution. Two drops of pyridine were added and the solution refluxed for 15 min., during which time maroon solid started precipitating. The mixture was cooled in a refrigerator and the solid was filtered and dried. The yield of I1 was 0.72 g. (99.5%). An analytical sample was prepared by two recrystallizations from 95% ethyl alcohol; m.p. 278-279' dec. The infrared spectrum of I1 showed absorption at 3180 (N-H), 2210 ( C s N ) , and 1672 cm.-' (C=O). In the ultraviolet spectrum in ethanol maxima were located a t 247 m r (log 6, 4.16), 281 l l l f l (log 6, 3.97), 345 mr (ah.) (log 6, 3.96), and365mp(loge, 3.98). Anal. Calcd. for CcH8NaOS: C, 53.73; H, 1.50; N , 20.89; S, 15.93; mol. wt., 201.2. Found: C, 53.44; H, 1.63; N, 20.86; S, 15.95; mol. wt., 194.8, 219.3. Isolation of Benzylidenemalononitrile (N).-The above procedure waa repeated using 2.0 g. of I and 1.84 g. of tetracyanoethylene. Immediately after the addition of the pyridine, 1.48 g. of benzaldehyde wa8 added. The precipitated solid (11) was filtered (2.28 g.). The filtrate was concentrated t o dryness under vacuum at room temperature. The residue was recrystallized from ethanol-water, filtered, and dried. The yield of IV was 1.05 g. (59% based on yield of I1 obtained). An analytical sample was prepared by two recrystalliEations from ethanol-water; m.p. 84-85'. A mixed melting point with an authentic sample of benzylidenemalononitrile showed no depression. The infrared spectra were also identical. Anal. Calcd.for CION~K;~: C,77.90; H , 3.92; N, 18.18. Found: C, 77.61; H, 3,96; N, 17.77. Diels-Alder Adduct of I1 (VII).--4 solution of 0.6 g. of 11, 0.02 g. of hydroquinone, and 4.5 ml. of 2,3-dimethylbutadiene in 30 ml. of ethyl alcohol was refluxed for 48 hr. The solvent WM removed under vacuum on a hot water bath. The residue ww twice recryatallized from ethanol-water. The yield of VI1 was 0.47 g. (55.7%). The white needles were recrystallized once more from ethanol-water to give an analytical sample; m.p. 220-221 '. Anal. Calcd. for C16H1sNsOS: C, 63.59; H, 4.62; K, 14.83. Found: C, 63.62; H,4.46; N, 14.84, 14.76. 2-Dicyanomethylene-2H,JH-benzo [b] thiophene-3-one (VI).-To a boiling solution of tetracyanoethylene (0.92 g.) in 45 ml. of benzene was added 0.93 g. of 2H,SH-benzo[b]thiophene-3-one (V)ll in 10 ml. of benzene. The mixture waa heated for 10 min. forming a deep maroon solution. Four drops of pyridine were added and the solution waa refluxed for 15 min. After cooling, the solid was filtered and dried. The yield of VI was 1.2 g. (91.7%). An analytical sample

(9) All melting points are uncorrected. Microanalyses and molecular weight determinations were performed by Mr. Josef Nemeth and his associates. N.m.r. speotra were measured b y 0. W. Norton and his associates on a Vsrian Associates A-60 spectrometer. Infrared spectra were measured b y D. H. Johnson and his associates on a Perkin-Elmer Model 2 1 spectrophotometer. (10) T. L. Cairns, R. -4. Carboni, D. D. Coffman, V. A. Engelhardt, R . E. Heckert, E. L. Little, E. 0 . McGeer, €3. C. McKusick, W. J. Middleton, R. M. Scribner, C. W. Theobald, and H. -4. Winbere. J . Am. Chem. Sac., 80, 2775 (1958); R. A. Carboni, OTQ.Sgn., 99, 64 (1959). (11) C. Hansoh and H. G.Lindwall. J . 0 7 0 . Chem., 10,381 (194.5).

3890

SHAW

wm obtained by one recrystallization from 95% ethanol; m.p. 263-264" dec. The infrared spectrum of VI showed absorption a t 2220 (C=N) and 1685 em.-' (C=O). Anal. Calcd. for CIIHdN~OS:C, 62.25; H, 1.89; Y, 13.20; S, 15.10. Found: C, 62.15; H, 1.81; N , 13.04; S, 15.49. Diels-Alder Adduct of VI (VIII).-A solution of 0.5 g. of VI, 0.02 g. of hydroquinone, and 4.5 ml. of 2,a-dimethyl-

VOL. 27

butadiene in 50 ml. of ethyl alcohol was refluxed for 48 hr. The solvent wm removed under vacuiim on a hot water bath. The residue was recrystallized from ethanol-water. The yield of VI11 was 0.65 g. (93.8%). The white needles were recrystallized twice from ethanol-water to give an analytical sample; m.p. 157-158". Anal. Calcd. for CnH14N20S: C, 69.35; H, 4.79; S , 9.52. Found: C,69.57; H,4.97; N,9.52.

The Preparation of +Triazine Derivatives Containing the N-0 Bond. I. Mono-N-oxides of Amino-substitu ted .+Triazine Derivatives1 JOHN T. SHAW Research Department, Organic Chemicals Division, American Cyanamid Company, Bound Brook, ,Vew Jersey Received April 9, 1961 The preparation of mono-N-oxides of amino-substituted s-triazine derivatives by two methods is reported: peracetic acid oxidation of aminotriazines and cyclization of dicyanoamidine salts with hydroxylamine hydrochloride. Both methods give the same N-oxide. The dicyanoamidine salts were obtained by reaction of alkyl imidate hydrochlorides with sodium acid cyanamide (limited to lower alkyl) or by reaction of alkyl- or arylamidines with cyanogen chloride. The latter method appears to be general.

The objective of this investigation was the prep- elemental analysis, and infrared comparisons with aration of mono-N-oxides of amino-substituted an authentic sample. s-triazine derivatives. A review of the literature With the previous data in mind, the structural uncovered only one reference to the preparation possibilities CaH&s0 appeared to be I, 11, or 111 of an s-triazine mono-N-oxide. I n this work, by CH3 Kaiser and Roemer,2 a good yield of melamine Noxide was obtained by slurrying an equimolar mixture of potassium dicyanoguanidine and hydroxylamine hydrochloride in Cellosolve. This cyclization method was not chosen initially primarily because the dicyanoamidines required in the reaction I1 were themselves unknown. An alternative route to the N-oxides was an oxidative procedure. and tautomers thereof. The normal triazine NGrundmann and Schroedera had shown that Carols oxide structures are shown for sake of simplicity. acid oxidation of 2-amino-4,6-bis-p-chlorophenyl- The other triazine isomer of C,H,NsO, 2,bdiaminos-triazine gave a tris-N-oxide. It was hoped that 6-hydroxymethyl-s-triazine (which was not really the use of a milder peracid like peracetic acid under expected), was eliminated on the basis of recovering the proper reaction conditions would lead to mono- acetoguanamine from the phosphorus trichloride hT-oxideformation. reaction. The Peracetic Acid Oxidation of Amino-substituted s-Triazines.-The reaction of acetoguanCH3 amine (1 mole) with peracetic acid (1.5 moles) was carried out in acetic acid a t 4045" for about twenty-four hours. The compound isolated was a white crystalline solid, whose elemental analysis gave the empirical formula C4H7Kt,0; no water I n the synthesis of 111, freshly prepared potaspresent. The material was soluble in dilute acid sium amide was added in the cold to 2,4-dichloroand base and gave a deep red color with ferric 6-methyl-s-triazine4 in ether, followed by the addichloride. To confirm the assumption that the tion of an intimate mixture of hydroxylamine triazine ring was still intact, the oxidized product hydrochloride and sodium carbonate. The usual was treated with phosphorus trichloride in chloro- physical properties of melting point. infrared, and form to deoxygenate it. Acetoguanamine was solubility of 111 did not agree with those of the identified as the resulting product by melting point, peracetic acidacetoguanamine oxidation product. Furthermore, 111 gave a blue color with ferric (1) Presented in part before the Division of Organic Chemistry a t chloride in contrast to the red color of the acethe 140thNational Meeting of the American Chemical Society, Chicago, toguanamine product. Ill., September, 1961. (2) D. W. Kaiser and J. J. Roemer. U. S. Patent 2,729,640 (January 3, 1956). (3) C. Grundmann and H. Schroeder, Chem. Bcr., 87, 747 (1954).

(4) R. Hirt, H. Nideaker, and R. Berobtold, Helu. Chim. Acta. 33, 1365 (1950).